Agricultural economists commonly base their agricultural price pohcy recommendations on costs of production (although wtth reservations). Their reseivat10ns are centred on the problems of assigning values to the contributions of inputs: durable and those supplied by the farm operators' household. Some of the reseivations can be reduced if, when developing the production cost estimates, the focus is placed on the policy goal to be achieved The policy goal affects both the selection of the population for which the cost estimates are made and the items to be included in the cost estimates. Policies designed to provide incentives to producers must be targeted towards the costs of those producers supplymg the maJOr portions of output of the commodities m question. The costs of an often relatively small number of large producers thus become heavily weighted. However, policies to sustain the mcome of selected groups of farmers must be based on the costs of production expenenced by a large number of small farmers. Cost estimates should distinguish between financial costs and economic costs. Structural differences exist in the cost structure among farmers. In setting prices, major attention should be given to the annual financial costs of capital replacement. Pnce policies cannot be directed to meet a specifically targeted economic cost